Kenji Doya1, Kazuyuki Samejima, Ken-ichi Katagiri
1Human Information Science Laboratories, ATR International, Seika, Soraku, Kyoto 619-0288, Japan. doya@atr.co.jp
You might also read
Articles linked to this work by shared authors, journal, and citation graph.
We introduce multiple model-based reinforcement learning (MMRL), a modular approach for complex control tasks. MMRL decomposes problems, enabling adaptive learning in changing environments for improved performance.
Area of Science:
Background:
Purpose of the Study:
Main Methods:
Main Results:
Conclusions: